MAJA PLANINIC, ZELJKA MILIN-SIPUS, HELENA KATIC, ANA SUSAC

and LANA IVANJEK

COMPARISON OF STUDENT UNDERSTANDING OF LINE GRAPH

SLOPE IN PHYSICS AND MATHEMATICS
Received: 12 August 2011; Accepted: 15 April 2012

ABSTRACT. This study gives an insight into the differences between student
understanding of line graph slope in the context of physics (kinematics) and
mathematics. Two pairs of parallel physics and mathematics questions that involved
estimation and interpretation of line graph slope were constructed and administered to 114
Croatian second year high school students (aged 15 to 16 years). Each pair of questions
referred to the same skill in different contexts—one question in the context of
mathematics and the other in the context of kinematics. A sample of Croatian physics
teachers (N = 90) was asked to rank the questions according to their expected difficulty for
second year high school students. The prevalent ranking order suggests that most physics
teachers expected mathematics questions to be more difficult for students than the parallel
physics questions. Contrary to the prevalent teachers’ expectations, students succeeded
better on mathematics than on physics questions. The analysis of student answers and
explanations suggests that the lack of mathematical knowledge is not the main reason for
student difficulties with graphs in kinematics. It appears that the interpretation of the
meaning of line graph slope in a physics context presents the largest problem for students.
However, students also showed problems with the understanding of the concept of slope
in a mathematical context. Students exhibited slope/height confusion in both contexts, but
much more frequently in the context of physics than in the context of mathematics.

KEY WORDS: kinematics, line graph, mathematics education, physics education, slope

INTRODUCTION

The fact that many students at high school, or even university level, lack the
ability to understand and interpret graphs in physics is not new. It has been
investigated in several physics education research studies (e.g. Arons, 1983;
Beichner, 1990, 1994; McDermott, Rosenquist & Zee, 1987; Wavering,
1989; Woolnough, 2000; Brassel & Rowe, 1993). This topic has also been
the subject of mathematics education research (Dreyfus & Eisenberg, 1990;
Graham & Sharp, 1999; Habre & Abboud, 2006; Kerslake, 1981;
Leinhardt, Zaslavsky & Stein, 1990; Swatton & Taylor, 1994), since

This research is a part of the scientific project 119-0091361-1027 funded by the Ministry of Science,
Education and Sports of the Republic of Croatia.

International Journal of Science and Mathematics Education (2012) 10: 1393Y1414

# National Science Council, Taiwan 2012
1394 M. PLANINIC ET AL.

line graphs were first introduced into mathematics and then only
later, into physics. Both disciplines require students to be able to
extract various pieces of information from line graphs, but, in
addition, physics also requires an interpretation of the obtained
information in the context of a given physical situation. One such
type of information that students are very often asked to estimate and
interpret is the slope of the line graph.
The concept of slope (gradient) is very important for physics since
many physical quantities are defined as gradients (e.g. velocity,
acceleration) and represented with line graphs. The concept of slope is
also important for mathematics as a necessary prerequisite for the
development of the concept of derivation. Students study line graph
slope in both mathematics and physics, but, because of differences in
contexts, they may not necessarily realize that they are studying the same
concept. Student difficulties with the concept of slope were identified
through both physics and mathematics education research, usually as a
part of studies which investigated general student difficulties with graphs.
The study of McDermott et al. (1987) presented a good overview of
student difficulties with graphs. Regarding slope, it was found that
students have difficulties discriminating the slope and height of a graph
and interpreting changes in height and changes in slope. Students often do
not know whether to extract the desired information from the slope or the
height of the graph, and instead of looking for changes in slope, many
students focus on the more perceptually obvious changes in height.
On the basis of reports on student difficulties with graphs, Beichner
(1994) constructed the Test of Understanding Graphs in Kinematics
(TUG-K) and applied it to 895 high-school- and college-level students.
He pointed out that the most common mistakes students make with
kinematics graphs are thinking that a graph is a picture of the situation
and confusing the meaning of the slope of the line with the height of a
point on the line. This study also stressed that many students were unable
to choose which feature of the graph represented the information that was
needed to answer the question (for example, they calculated slope when
they should have been calculating the area). Graphing skills of high
school students were investigated in a study (Brassel & Rowe, 1993),
which found that at least one fifth of the students did not have adequate
graphing skills. Students had difficulties with linking the graph and the
verbal descriptions of a given event, and they did not understand graphs
as a means of representing relationships among variables. Students’
facility with graphs was found to be generally superficial, being based on
a few simplistic algorithms.
 COMPARISON OF STUDENT UNDERSTANDING OF LINE GRAPH SLOPE 1395

The use of microcomputer-based labs and real-time graphing was

investigated and found to have an effect, especially on reducing “graph as
picture” errors (e.g. Mokros & Tinker, 1987).
Most of the research on student understanding of graphs in physics was
done in the context of kinematics, because of the very broad use of
graphical representations in kinematics. As well, some of the research in
mathematics education was also based on kinematics motion graphs
(Graham & Sharp, 1999; Kerslake, 1981). General findings are similar to
those in physics education research: Students tend to use graphs as actual
pictures of the motion. They do not realize which feature of the graph to
use in a particular situation, and they tend to use a position criterion
instead of a gradient-based criterion when considering velocities. One of
the studies (Graham & Sharp, 1999) focused only on able 13-year-old
students with interest in mathematics and found that relatively few of
them had a sound understanding of graphs that could be applied
consistently.
The study of Woolnough (2000) was particularly interesting because it
revealed the existence of student resistance to applying their mathematical
knowledge to physics. The study suggested that senior secondary students
operated in three distinct contexts: the real world, the physics world, and
the mathematical world, each with different characteristics and belief
systems. It was suggested that, most students, even those who do well in
mathematics and physics, do not make substantial links between these
contexts. Some students even thought that it was not appropriate to
transfer concepts from mathematics to physics. For example, when
calculating the slope of a line graph, some students thought that it was
inappropriate to assign units to the slope because of their perception that
slope is a mathematical concept. The same study also found that about
half of the high school students entering Year 11 physics were not overly
familiar with the concept of slope and could not determine the slope of a
simple straight line. In Year 12, the situation greatly improved, and it was
found that 90% of students were able to calculate a slope of a line through
origin, but almost a third of those students were not able to provide a
physical interpretation of the calculated slope.
Studies in a purely mathematical context (e.g. Dreyfus & Eisenberg,
1990; Habre & Abboud, 2006; Leinhardt et al. 1990; Swatton & Taylor,
1994) have shown that student understanding of mathematical concepts
(such as functions) tends to be typically algebraic and not visual. Visual
information is more difficult for students to learn and is considered less
mathematical (Dreyfus & Eisenberg, 1990; Habre & Abboud, 2006).
Leinhardt et al. (1990) classified student difficulties into three categories:
1396 M. PLANINIC ET AL.

interval/point confusions, where students focus on a single point instead

of on an interval; slope/height confusions where students mistake the
height of the graph for the slope; and iconic confusions, where students
incorrectly interpret graphs as pictures.
It appears that the results of investigations of student difficulties with
graphs in physics and mathematics have a lot in common. The same
difficulties are found in both contexts—the slope/height confusion and
iconic confusion seeming to be the most prominent.

Theoretical Background
The first step in extracting any information from a graph is realizing that
it is a symbolic representation of the relationship between variables.
Processing of visual symbolic information, such as line graphs, requires
the ability to perceive and remember a pattern of spatially arranged visual
data as well as the ability to reason about spatial visual information. It is
therefore not surprising that understanding of kinematics graphs, and in
particular, slope calculation, was found to be related to logical thinking,
spatial ability, and mathematics achievement (Bektasli, 2006).
How graphs are related to student cognitive development was
investigated in a study (Wavering, 1989) which found a correspondence
between categories in student graphing skills and Piagetian stages of
cognitive development. This study and other similar ones (Berg &
Phillips, 1994) suggest that constructing and interpreting graphs requires
formal operational reasoning. Students who had low levels of logical
thinking were not able to construct and interpret graphs. Berg & Phillips
(1994) state that, without developed logical thinking, students are
dependent upon their perceptions and low-level thinking, which will lead
them to a “graph as picture” error, slope/height confusion, etc.
Students who have not yet reached the formal operational stage of
cognitive development are likely to view graphs as something concrete
rather than as indicators of abstract trends (Beichner, 1990). This view
was criticized by Roth & McGinn (1997) who emphasized graphing as
practice (as opposed to cognitive ability) and attributed student difficulties
with graphing to students’ lack of experience with the construction and
use of graphs, as well as to a lack of opportunities for students to endow
graphs with meaning.
Spatial ability can be defined as the intuition about shapes and the
relationships among shapes, that is, as the ability to generate, retain,
retrieve, and transform well-structured visual (mental) images (Lohman,
1996). Researchers in cognitive psychology have suggested that measures
 COMPARISON OF STUDENT UNDERSTANDING OF LINE GRAPH SLOPE 1397

on spatial ability tests reflect both subjects’ ability to maintain and

transform spatial images, as well as the capacity of their visual–spatial
working memory (Miyake, Friedman, Rettinger, Shah & Hegarty, 1991;
Salthouse, Babcock, Mitchell, Palmon & Skovronek, 1990; Shah &
Miyake, 1996). It was found that people who differ in spatial abilities also
differ in their ability to solve physics problems that involve multiple
spatial parameters (Kozhevnikov, Motes & Hegarty, 2007). Interpretation
of kinematics graphs is an example of such problems. Vekiri (2002)
found that students with low prior knowledge and low spatial ability have
difficulties extracting information from graphs.
Kozhevnikov, Hegarty & Mayer (1999) investigated student graph
interpretation in correlation with visual ability (the ability to construct
mental images of vivid color and detail) and spatial ability. Visual ability
refers to physical objects in the real world and is therefore more related to
concrete operations, whereas spatial ability refers to three-dimensional
transformations of these objects and is more related to formal operations.
The same study searched for a relationship between imagery and
problem solving in kinematics. Two different types of imagery were
defined, visual and spatial. Visual imagery refers to the external
appearance of objects (e.g. color, shape). Spatial imagery refers to spatial
relationships between parts of an object and the location of objects in
space. It was found that spatial imagery contributes to student graph
interpretation in physics, whereas visual imagery is an obstacle on the
same task. The authors selected kinematics as the context for their study
because this topic is related to visual and spatial abilities since it involves
both graphical and concrete physical representations. The authors argue
that focusing on concrete aspects of the situation prevents students from
thinking formally. Students who are more visual (low spatial ability), who
generated vivid images of the concrete situation (visual imagery), tended
to interpret the graph as a literal picture of the situation and were unable
to abstract any relevant information from the graph. Students who are
more spatial, on the other hand, considered the graph to be an abstract
spatial representation, and none of them referred to the graph as a
concrete picture of the motion. Students of low spatial ability are
therefore expected to have more problems with graph interpretation than
high-spatial-ability students.
Another important factor in graph interpretation is students’ conceptual
knowledge in physics. It appears that a correlation between spatial ability
and kinematics problem solving is no longer present after students receive
physics instruction through the use of rich visualization technologies
(Kozhevnikov & Thornton, 2006). Spatial ability and physics conceptual
1398 M. PLANINIC ET AL.

knowledge are probably interrelated since high spatial ability may

enhance people’s ability to gain a conceptual knowledge of physics
(Kozhevnikov et al., 2007). It is suggested that a physics curriculum that
provides external visualizations can help students who have difficulty
generating such visualizations on their own (low spatial ability students),
to compensate for such shortcomings. In the case of kinematics graphs,
visual simulations should highlight segments of data and associated tick-
mark ranges along the axes, rather than the overall shape of the graph, and
should lead the learner to analyze and imagine event changes occurring
within the subintervals (Kozhevnikov et al., 2007).
To summarize, most cognitive studies cited above point to the
importance of logical thinking and spatial ability as prerequisites for
understanding kinematic graphs, and some also suggest that studying
kinematics graphs in physics and mathematics can help develop these
abilities in students.

Physics Teachers’ Attitudes

Physics teachers are often inclined to attribute the observed student
difficulties with graphs to students’ lack of mathematical knowledge.
However, some researchers in physics education (McDermott et al., 1987)
have suggested that the lack of mathematical skills may not be the main
cause of students’ difficulties with graphs in physics. Also, some
researchers in mathematics education have pointed out that often students
who solve graphing or function problems in mathematics seem to be
unable to access their knowledge in science (Leinhardt et al., 1990). To
further investigate this issue, we have undertaken a study to investigate
Croatian students’ understanding of line graph slope in mathematics and
kinematics contexts, as the first step in the development of an instrument
which could measure and compare student understanding of graphs in the
broader contexts of physics and mathematics. We have also surveyed a
group of Croatian physics teachers (N = 90) and asked them to rank
parallel physics and mathematics questions which are concerned with the
concept of slope, according to their expected difficulties for students.
Teachers’ predictions were compared with the empirically obtained
difficulties of questions.

Research Questions
This study attempts to answer the following research question: How does
student ability to estimate and interpret slopes of line graphs in
mathematics relate to their ability to estimate and interpret slopes of line
 COMPARISON OF STUDENT UNDERSTANDING OF LINE GRAPH SLOPE 1399

graphs in physics, e.g. kinematics? Although a number of studies

investigated these abilities separately, we are not aware of the existence
of a study which attempted to compare them directly on parallel physics
and mathematics problems. Such an attempt could provide new insights
into the relationship between student understanding of line graph slope in
mathematics and physics and the frequency of typical student errors in
each context. An additional purpose of the study is to document and
compare Croatian student difficulties concerning line graphs, with the
corresponding documented difficulties of students in other countries.

METHODOLOGY

Two pairs of parallel multiple choice mathematics and physics questions

about graphs were developed by authors (Figures 1 and 2). Each pair of
questions referred to the concept of line graph slope in different contexts
—one question in a mathematics context and the other in a physics
context. The first pair of questions referred to positive slope, and the next
pair to negative slope. Physics questions (labeled P1, P2) were adjusted
questions from Beichner’s TUG-K (Beichner, 1994), and mathematics
questions (labeled M1, M2) were constructed to be analogous to the

Figure 1. Questions P1 and M1

1400 M. PLANINIC ET AL.

Figure 2. Questions P2 and M2

respective physics questions. The mathematics and physics questions are

regarded in this study as parallel (M1 to P1 and M2 to P2) in the sense
that the required mathematical reasoning in both corresponding questions
is the same.
In question P1, the correct answer is given in option B and, in question
M1, in option A. In both questions, students were expected to reason
about the slope of the line graph in order to recognize the slope as
constant and different from zero and in question P1, to also interpret the
slope as the magnitude of the object’s velocity. In question P2, the correct
answer is given in option A and, in question M2, in option B. In these two
questions, students were expected to recognize that the slope of the graph
was constant and negative. In Question P2, they were also expected to
interpret the slope of the graph as the magnitude of the object’s
acceleration.
All four questions were administered to 114 second year high school
students (aged 15 to 16 years) from two Croatian cities, Zagreb and Slunj,
as a part of a longer (16 questions) multiple choice test about graphs.
Parallel questions analyzed in this study did not follow each other in the
test but were separated with other questions. The test was administered in
written form, and the allocated time for taking the whole test was 45 min.
One of the researchers (H. K.) was always present at the time of the
 COMPARISON OF STUDENT UNDERSTANDING OF LINE GRAPH SLOPE 1401

testing to answer students’ questions and supervise the testing. There was
no incentive such as grades offered for taking the test. Students were
informed that the test was a part of the research on student understanding
of graphs, and they were generally willing to participate. In addition to
choosing the correct answer, students were asked to provide explanations
for their answers so that an insight into the underlying student reasoning
could be obtained. Unfortunately, not all students provided explanations
for their answers. In cases where there was no explanation, the answer
was taken at its face value.
There were 52% male students and 48% female students in the sample.
Most of the sample was from the Croatian capital Zagreb (80 students),
and 34 students were from the smaller town of Slunj. In the first year of
high school, all students have studied, among other topics, linear
functions and linear graphs in mathematics and motion graphs and
kinematics in physics.
Physics is a compulsory school subject in most Croatian high schools
which can be of different types. Sixty students were from high schools
which specialize in natural sciences and mathematics (NSM); 33 students
were from high schools which are of a general education type (GE), and
there were 21 students from a vocational high school for informatics
(VS). Students in NSM schools had three class periods of physics and
four periods of mathematics per week, whereas students in GE schools
had two periods of physics and four periods of mathematics per week.
Students in VS schools had two periods of physics and three periods of
mathematics per week in the first year of high school.
Regarding abilities in mathematics and physics, the students from
NSM schools are from the top 20% of the Croatian high school students;
GE students can be considered as being average and VS students slightly
below average.
Physics teachers, who gathered in April 2011, at the biannual Croatian
Symposium on Physics Teaching, were surveyed about their expectations
of the relative difficulties of the four questions P1, P2, M1, and M2. In
the written questionnaire, teachers were asked to rank these four questions
according to their expected difficulty for second year high school
students, starting with the question with the highest level of difficulty.
They based question rankings on their experience and opinion. Teachers
were also asked to provide reasons for the ranking order that they
proposed. Ninety teachers filled out the questionnaire. Their answers were
analyzed and compared with the actual level of difficulties established for
the four questions. Frequencies of different ranking combinations were
determined, and the explanations accompanying these combinations were
1402 M. PLANINIC ET AL.

grouped according to similarities in teachers’ reasoning. In addition, the

answers of teachers who teach both mathematics and physics were
analyzed separately and compared with the answers of other teachers who
teach physics, but not mathematics.

RESULTS AND DISCUSSION

Table 1 gives an overview of student average success on each question

for different types of schools. It is evident that students in all types of
schools scored higher on mathematics questions than on the
corresponding physics questions. This is not surprising because, in
addition to mathematical skills, the physics questions also required
students to interpret mathematical results in the context of the physics
situation. The overall success on the four questions was relatively high in
NSM and GE schools but much lower in VS. This is not only the
consequence of a larger number of physics and mathematics lessons in
NSM and GE schools, but also of higher motivation and higher general
ability of students in those schools. In Table 2, the frequencies of different
combinations of correct and incorrect answers for questions P1, P2, M1,
and M2 are given.
In Figure 3, percentages of correct answers for pairs of parallel physics
and mathematics questions are shown. It is immediately noticeable that
parallel physics and mathematics questions differ in level of difficulty.
Also, the results on questions P1 and P2 are quite close to the results of
university students in Beichner’s study (1994) on the same questions,
where there was 63% of correct answers for question P1 and 18% of
correct answers for question P2.

TABLE 1
Number (N) and average success of students from different types of schools (NSM, GE,
VS) on physics (P1, P2) and mathematics (M1, M2) questions

Type of school N P1 P2 M1 M2

NSM 60 71.7% 33.3% 91.7% 80.0%

GE 33 57.6% 12.1% 60.6% 48.5%
VS 21 28.6% 14.3% 52.4% 19.0%
All schools 114 59.6% 23.7% 75.4% 59.6%
 COMPARISON OF STUDENT UNDERSTANDING OF LINE GRAPH SLOPE 1403

TABLE 2
Number (N) and percentage of students for different combinations of correct and incorrect
answers on questions P1, P2, M1, and M2

P1 and M1 P2 and M2 P1 correct, P1 incorrect, P2 correct, P2 incorrect,

correct correct M1 incorrect M1 correct M2 incorrect M2 correct

N 56 24 12 30 3 44
% 49, 1 21, 1 10, 5 26, 3 2, 6 38, 6

Positive Slope
Questions M1 and P1 both referred to the positive slope of a line graph.
Both questions are shown in Figure 1. Mathematics question M1 was
easier for students than the physics question P1. In question M1, the word
‘slope’ was already mentioned in the text of the question, whereas in
question P1, students had to determine on their own that the slope of the
graph was the decisive feature for answering the question about the
motion of the body. They also had to know that the slope of s vs. t graph
represents the magnitude of the body’s velocity. This obviously increased
the difficulty of question P1 compared with question M1. In some
explanations given by students with the correct answer to question P1,

Figure 3. Percentages of correct answers for pairs of parallel physics and mathematics
questions
1404 M. PLANINIC ET AL.

neither the word slope nor the idea of the slope was present. Here are
some examples:
The ratio of the covered distance and the time (the speed) is always the same.
Because the covered distance in each moment increases by the same amount.

In the second comment, the student does not distinguish a moment

from a time interval, but they probably understand (at least vaguely) the
idea of slope.
From some comments, it was hard to see what the student was
thinking. An example of such a comment is the following one given with
the correct answer to question P1:

s=vt, v=const.

Since the student had chosen the correct answer, he might have
indicated with this equation his understanding of the velocity as the slope
in s vs. t graph, but we cannot be completely sure of this.
Some students used the concept of slope in their reasoning, as can be
deduced from the following comment:

The velocity is constant, since s vs. t graph is a straight line with the constant slope.

Some explanations that accompanied correct answers to question M1

revealed that these students’ understanding of the concept of slope is at
best only partially correct, or even completely wrong:

Because the slope is the angle between the straight line and the x axis.
The slope of the straight line is not changing and the line is all the time in the first
quadrant.
The slope angle is not changing and it is greater than zero.
The straight line does not begin from zero, therefore it (the slope) is different from zero
and the line is in the positive quadrant.

It can be inferred from the explanations above that some students tend
to attribute zero value of slope to the straight line that passes through the
origin of the coordinate system, while others associate slope value with
the quadrant in which the line is drawn (not realizing that the straight line
extends to infinity on both sides and has only one value of slope). Some
identify slope with the angle between the straight line and the x axis.
The leading and largely prevailing wrong answer chosen by students in
P1 (40%) was the answer D (the body moves with uniformly increasing
velocity), and in M1, it was the answer C (23%) (the slope of the line
constantly increases). The most frequent wrong choice on the question
 COMPARISON OF STUDENT UNDERSTANDING OF LINE GRAPH SLOPE 1405

similar to P1 in Beichner’s study (1994) was also the one that suggested
that the body moves with uniformly increasing velocity. In both contexts,
physics and mathematics, students most often made the same well-known
mistake, which consisted in confusing the slope of the graph with the
height of the graph (Beichner, 1994; McDermott et al., 1987; Leinhardt et
al., 1990). However, the frequency of this mistake was much higher in the
context of physics than in the context of mathematics.
One student showed in his answer to M1, an even more profound
confusion:
The correct answer is C (the slope constantly increases) because the body’s motion is
uniformly accelerated.

This student confused the mathematics graph with the v vs. t graph,
which could be a sign of a known student tendency to memorize shapes
of kinematics graphs without paying attention to what physical quantities
(if any) are depicted in the graph (Beichner, 1994).
Most of the students who answered the physics question P1 correctly
also succeeded on the parallel mathematics question (about half of the
students got both questions correct). It is interesting, however, that about
26% of students (Table 2) correctly answered the mathematics question
but failed on the parallel physics question. (The opposite was the case in
about 10% of students. Some students could have solved P1 without
understanding the slope concept if they, for example, memorized the
shape of the s vs. t graph for motion at constant velocity). This indicates
that mathematical knowledge is not a guarantee of success on a parallel
physics problem. Problems with student understanding of the concept of
line graph slope in mathematics were also observed. We see from
explanations which accompanied answers to M1 that student mathemat-
ical understanding of the concept of slope was often only partially correct,
even in cases when students chose the correct answer.

Negative Slope
Questions P2 and M2 are shown in Figure 2. They are both harder than
P1 and M1, respectively (Figure 3, Tables 1 and 2). Negative slope is
obviously a harder concept than positive slope, but this is even more
pronounced for the concepts of acceleration and velocity.
Some students showed elements of correct understanding in their
explanation for question P2, such as the student who gave the following
explanation:
1406 M. PLANINIC ET AL.

A is the correct answer, because in each moment velocity diminishes by the same amount.

Once again we notice that some students do not distinguish a moment

of time from an interval of time, but the underlying reasoning about
acceleration is essentially correct.
Most students did not answer question P2 correctly. The following
explanations which accompanied answer D (acceleration decreases
uniformly), represent incorrect reasoning based on poor understanding
of uniformly decelerating motion.

In uniformly decelerated motion velocity decreases linearly, v=at.

The answer is D because the body decelerates uniformly.

Students notice correctly that the depicted motion is uniformly

decelerated (velocity diminishes at a steady rate), but then incorrectly
link this type of motion with uniformly decreasing acceleration. This
indicates that they do not understand the concept of acceleration either as
the rate of change of velocity, or as the slope of the line in the v vs. t
graph.
There were more correct answers to question M2 than to question P2,
but not many explanations were given for either of them. Here are some
explanations for the correct answer to question M2:

The graph is a straight line and it has negative direction.

The angle between the straight line and the x axis is the same and the value of the function
is decreasing.

The first explanation suggests that the student’s reasoning is based on

the appearance of the graph and probably also on the learned rule that the
slope of the line is negative when the line has negative direction (negative
direction meaning that y values become smaller as x values increase). It is
not possible to judge whether this student really understands the concept
of slope. In the second explanation, we see an indication of the correct
reasoning based on the concept of slope, but we can also notice once
again the possible identification of slope with the angle between the
straight line and the x axis.
For both questions P2 and M2, the leading wrong answer was the same
and indicated the presence of the well-known slope/height confusion: The
acceleration of the body (in P2) or the slope of the graph (in M2)
constantly decreases. The same was the case in Beichner’s study (1994)
for question P2, where 72% of students chose a similar answer. However,
the frequency of the same mistake in our study was again very different in
the contexts of physics and mathematics. In the physics context, 68% of
 COMPARISON OF STUDENT UNDERSTANDING OF LINE GRAPH SLOPE 1407

students displayed slope/height confusion (even though some of them

correctly stated in their explanations that the body uniformly decelerated)
whereas, in mathematics context, only 33% of the students were similarly
confused. Almost 40% of students correctly answered the mathematics
question but failed on the parallel physics question, whereas the opposite
was true for less than 3% of students (Table 2). This suggests that the
interpretation of mathematical quantities in the context of physics is an
important source of student difficulties with graphs in physics. Many
students who are able to estimate whether the slope of a straight line is
constant (as is visible from their answers to question M2) fail to conclude
that the acceleration is constant in P2. This is obviously not the
consequence of their lack of mathematical knowledge but rather of the
missing link between mathematics and physics and of the lack of relevant
conceptual physics knowledge. Students are probably not aware that the
problem in P2 is mathematically the same as the problem in M2, because
they do not see acceleration as the slope of the v vs. t graph. Also, student
understanding of the concept of acceleration is obviously problematic—
some associating finite change in velocity with an instant of time instead
with a finite interval of time (interval/point confusion) and many thinking
that uniform deceleration implies a decreasing acceleration.
The difference in student performance on parallel mathematics and
physics questions may also be explained in terms of the different
cognitive demands that these problems impose on students. The cognitive
literature suggests that generating visual imagery when interpreting
graphs is an obstacle to problem solving in physics. Mathematics
problems are less likely to induce visual imagery than physics problems,
which contain more “real-life” associations (object, motion, speed, etc.).
The context of physics problems could have led some students to start
thinking about what the object in question or its trajectory might look like
and then to connect that trajectory erroneously with the shape of the
graph, therefore failing to perceive the graph as a symbolic representation
of a relationship between variables. Physics questions are also less direct.
Mathematics questions directly called for the activation of knowledge
related to slope, and students correctly answered questions M1 and M2 if
they possessed such knowledge—even if that knowledge was in some
cases only partially correct. However, on physics questions P1 and P2,
many students did not even activate their mathematical knowledge about
slope, because they lacked relevant conceptual knowledge about velocity
and acceleration that should have led them to use this mathematical
knowledge in a physics context. Instead, they activated naive thinking
patterns under the influence of a strong visual cue (increasing/decreasing
1408 M. PLANINIC ET AL.

height of the graph). Different frequencies of naive answers in the two

contexts suggest that students used different strategies for the same type
of problems in mathematics and in physics. The added context in physics
problems made these problems more complex than the parallel mathe-
matics ones. In consequence, the results suggest that the physics problems
required more information processing from students than the respective
mathematics problems and were cognitively more demanding for them.

Survey of Physics Teachers

The results of the survey of physics teachers are shown in Figures 4, 5,
and 6. The observed leading rank order (from the most difficult to the
least difficult question) is M2–M1–P2–P1 (Figure 6), by which teachers
expressed their opinion that mathematics is more difficult for students
than physics. If we add up all combinations that have the both
mathematics questions before the both physics questions, we find that
33 out of 90 teachers (37%) shared this opinion. It is interesting that the
second most frequent ranking combination (P2–P1–M2–M1) is very close
to the order of difficulty found in our study (P2–P1=M2–M1), in which
P2 was the most difficult question, P1 and M2 were equally difficult, but
easier than P2, and M1 was the least difficult question. If combination
3412 (P2-M2-P1-M1), which also cannot be distinguished from the
empirically found ranking order, were added to the combination 3142
(P2–P1–M2–M1), we could say that 18 out of 90 teachers (20%) correctly
estimated the order of difficulty for the four questions. The same number
of teachers ranked physics questions as more difficult than mathematics
questions. The prevailing attitude among teachers, expressed through the
ranking of questions, was that each mathematics question would be

Figure 4. Percentage of teachers’ answers in which question M1 is ranked as more

difficult than question P1 and those in which P1 is ranked as more difficult than M1
 COMPARISON OF STUDENT UNDERSTANDING OF LINE GRAPH SLOPE 1409

Figure 5. Percentage of teachers’ answers in which question M2 is ranked as more

difficult than question P2 and those in which P2 is ranked as more difficult than M2

harder for students than the respective parallel physics question (Figures 4
and 5). In the explanations for the ranking order that they provided, many
teachers expressed the idea that the physics context is more familiar to
students than the mathematics context and that it is therefore easier for
students to understand graphs in physics than in mathematics. In other
words, many teachers consider mathematics more abstract and therefore
more difficult than physics. Some of the teachers also mentioned that one
of the leading causes of student difficulties with physics is students’ lack
of mathematical knowledge and skills. Twenty of the surveyed teachers
were both physics and mathematics teachers. They ranked P1 as more
difficult than M1 in 11 cases (55%) and P2 as more difficult than M2 in

Figure 6. Frequency of different ranking combinations given by physics teachers (‘1’

stands for question P1, ‘2’ for M1, ‘3’ for P2, and ‘4’ for M2)
1410 M. PLANINIC ET AL.

ten cases (50%), which is higher than in the whole sample. This could
indicate that teachers who teach both subjects have a somewhat better
insight into student difficulties, but, because of the small number of these
teachers, we cannot be sure that the observed differences are significant.
The study also provided some information on Croatian high school
students. The same most common difficulties with the concept of line
graph slope that were found in students from other countries (e.g. Araujo,
Veit & Moreira, 2008; Cataloglu, 2007) are also found in this sample of
Croatian students. Comparison of student success on questions P1 and P2
with the success of American students in Beichner’s study (1994) shows
very similar results. This might suggest that typical student difficulties
with line graphs such as slope/height confusion reflect intuitive reasoning
patterns that are common for all students.
The results of Croatian students suggest that the intensity of student
difficulties varies with the type of school, but a larger sample would be
required to investigate the differences between schools in detail.
However, for students from all types of schools, the physics part of the
test was more difficult than the mathematics part, and even the most able
students (from NSM schools) had difficulties with interpretation of slopes
in a physics context.

CONCLUSION

Student understanding of line graph slope is important for both physics

and mathematics education. In this study, we have tried to compare
student ability to estimate and interpret the slope of a line graph in both
contexts by comparing student answers on two pairs of parallel
mathematics and physics (kinematics) questions. Many physics teachers
still believe that students’ lack of mathematical knowledge is the main
reason that students have difficulties with physics, as was confirmed also
by the results of our survey of physics teachers. The results of our study
seem to support the suggestion of McDermott et al., (1987) that the lack
of mathematical skills is not the main cause of student difficulties with
graphs in physics. Quite a number of students in our study were able to
solve a mathematical problem concerning slope, but failed on a parallel
physics problem. It is important to emphasize that some students really do
lack mathematical skills, but even if they do not, mathematical knowledge
is not a guarantee for their success on similar physics problems. Often,
students did not even recognize that the problems were similar—they
used different strategies for analyzing graphs in mathematics and physics.
 COMPARISON OF STUDENT UNDERSTANDING OF LINE GRAPH SLOPE 1411

The component of interpretation of mathematical quantities in the physics

context (e.g. recognizing the slope of a velocity—time graph as
acceleration) is often missing, and even if students possess the needed
mathematical knowledge, they will not use it in a physics context if they
cannot transfer understandings between a physics and mathematics
situation.
The transfer of knowledge from mathematics to physics seems to be
relatively weak, but that might be caused by the fact that students do not
see the similarity between certain problems in mathematics and physics.
Theories of transfer of knowledge are based upon the idea that knowledge
can be transferred from one situation to another and linked with a new
situation (Potgieter, Harding & Engelbrecht, 2008). Some researchers
disagree and argue that learners’ mental processes are structured by the
context and the implemented activities and tools (Lave, 1988). Potgieter
et al. (2008) suggest that teachers often expect students to rise above
context, but that it is not easy for students to apply the mathematics in
other contexts. Recognizing mathematics in a different context requires
good understanding of the context (which is often missing), along with
mathematical knowledge. This was also the case with students in this
study.
It seems that student knowledge is very compartmentalized and that
stronger links should be established between mathematics and physics
during teaching, but students also need stronger conceptual knowledge in
physics. However, we have also noticed from students’ explanations that
some of them already have significant problems with the concept of slope
in a purely mathematical context. Many identify slope with the angle
between the straight line and the x axis, or evaluate the sign of the slope
according to the quadrant in which the line is drawn. Several cases of
interval/point confusion were also found.
For learning and transfer of knowledge to succeed, both physics and
mathematics should focus more on the meaning (interpretation) of graphs.
It is very obvious from the analysis of student difficulties with graphs that
finding the meaning of graphs and of their characteristics (such as slope)
seems to be one of the most problematic aspects of graph analysis for
students. Students need to make sense of graphs during instruction;
discover and discuss their meaning through collaboration with their peers;
and if possible, connect graphs with real examples of motion (for example
through the use of motion detectors).
The findings of our study suggest that—contrary to the prevailing
expectations of physics teachers—physics problems concerning the slope
of a line graph are more difficult for students than parallel mathematics
1412 M. PLANINIC ET AL.

problems. Many physics teachers in our study thought that mathematics

items would be more difficult because they are more abstract and lack
context, whereas physics items, which are closer to real life and appear
less abstract, should be easier for students. The results suggest quite the
opposite. Even though mathematics questions appear more abstract, they
are more direct and require less processing of information and less
conceptual understanding than parallel physics questions. This study
suggests that it is the added physics context that makes parallel physics
problems more difficult, “masks” the mathematical essence of the
problem, and increases the cognitive demand on students. Teachers
should realize that it is very important to work on student conceptual
understanding and interpretation of physical and mathematical quantities,
as well as on building stronger links between the two subjects.
Another interesting result of this study is that the same common
student difficulty known from other studies as slope/height confusion is
found to be dominant in both contexts but is much more frequent in the
context of physics than in the context of mathematics. It also occurs more
frequently in questions concerning negative slope than in questions
concerning positive slope. A graph as a visual cue seems to provoke the
same naïve thinking in both contexts but not with equal frequency. In
mathematics, this kind of reasoning will be characteristic for students who
lack knowledge about slope. In physics, mathematical knowledge about
slope will often not be activated at all (even if it is present), because of a
lack of students conceptual knowledge. More students will therefore tend
to revert to their intuitive reasoning patterns on physics problems than on
similar mathematics problems.

ACKNOWLEDGMENTS

This research is a part of the scientific project 119-0091361-1027 funded

by the Ministry of Science, Education and Sports of the Republic of
Croatia.

REFERENCES

Araujo, I. S., Veit, E. A. & Moreira, M. A. (2008). Physics students’ performance using
computational modeling activities to improve kinematics graphs interpretation. Com-
puters in Education, 50, 1128–1140.
Arons, A. B. (1983). Student patterns of thinking and reasoning, part one. Physics
Teacher, 21, 576–581.
 COMPARISON OF STUDENT UNDERSTANDING OF LINE GRAPH SLOPE 1413

Beichner, R. J. (1990). The effect of simultaneous motion presentation and graph

generation in a kinematics lab. Journal of Research in Science Teaching, 27, 803–815.
Beichner, R. J. (1994). Testing student interpretation of kinematics graphs. American
Journal of Physics, 62, 750–762.
Bektasli, B. (2006). The relationships between spatial ability, logical thinking,
mathematics performance and kinematics graph interpretation skills of 12th
grade physics students. Unpublished doctoral dissertation, The Ohio State
University.
Berg, C. A. & Phillips, D. G. (1994). An investigation of the relationship between logical
thinking structures and the ability to construct and interpret line graphs. Journal of
Research in Science Teaching, 31, 323–344.
Brasell, H. M. & Rowe, B. M. (1993). Graphing skills among high school physics
students. School Science and Mathematics, 93(2), 63–69.
Cataloglu, E. (2007). Internet-mediated assessment portal as a pedagogical learning tool:
A case study on understanding kinematics graphs. European Journal of Physics, 28,
767–776.
Dreyfus, T. & Eisenberg, T. (1990). On difficulties with diagrams: Theoretical issues. In
G. Booker, P. Cobb & T. N. De Mendicuti (Eds.), Proceedings of the Fourteenth
Annual Conference of the International Group for the Psychology of Mathematics
Education (Vol. 1, pp. 27–36). Oaxtepex: PME.
Graham, T. & Sharp, J. (1999). An investigation into able students’ understanding of
motion graphs. Teaching Mathematics and its Applications, 18, 128–135.
Habre, S. & Abboud, M. (2006). Students’ conceptual understanding of a function and its
derivative in an experimental calculus course. The Journal of Mathematical Behavior,
25, 57–72.
Kerslake, D. (1981). Graphs. In K. M. Hart (Ed.), Children’s understanding of
mathematics (Vol. 11–16, pp. 120–136). London: John Murray.
Kozhevnikov, M., Hegarty, M., & Mayer, R. (1999). Students’ use of imagery in solving
qualitative problems in kinematics. Retrieved December 6, 2011, from http://www.
compadre.org/portal/services/detail.cfm?ID=2760.
Kozhevnikov, M., Motes, M. & Hegarty, M. (2007). Spatial visualization in physics
problem solving. Cognitive Sciences, 31, 549–579.
Kozhevnikov, M. & Thornton, R. (2006). Real-time data display, spatial visualization
ability, and learning force and motion concepts. Journal of Science Education and
Technology, 15, 113–134.
Lave, J. (1988). Cognition in practice: Mind, mathematics and culture in everyday life.
Cambridge: Cambridge University Press.
Leinhardt, G., Zaslavsky, O. & Stein, M. K. (1990). Functions, graphs, and graphing:
Tasks, learning, and teaching. Review of Educational Research, 60, 1–64.
Lohman, D. F. (1996). Spatial Ability and G. In I. Dennis & P. Tapsfield (Eds.), Human
abilities: Their nature and assessment (pp. 97–116). Hillsdale: Erlbaum.
McDermott, L. C., Rosenquist, M. L. & van Zee, E. H. (1987). Student difficulties in
connecting graphs and physics: Examples from kinematics. American Journal of
Physics, 55, 503–513.
Miyake, A., Friedman, N. P., Rettinger, D. A., Shah, P. & Hegarty, M. (1991). How
are visuospatial working memory, executive functioning, and spatial abilities
related? A latent-variable analysis. Journal of Experimental Psychology. General,
130, 621–640.
1414 M. PLANINIC ET AL.

Mokros, J. R. & Tinker, R. F. (1987). The impact of microcomputer-based labs on

children’s ability to interpret graphs. Journal of Research in Science Teaching, 24, 369–
383.
Potgieter, M., Harding, A. & Engelbrecht, J. (2008). Transfer of algebraic and graphical
thinking between mathematics and chemistry. Journal of Research in Science Teaching,
45(2), 197–218.
Roth, W. M. & McGinn, M. (1997). Graphing: Cognitive ability or practice? Science
Education, 81, 91–106.
Salthouse, T. A., Babcock, R. L., Mitchell, D. R. D., Palmon, R. & Skovronek, E. (1990).
Sources of individual differences in spatial visualization ability. Intelligence, 14, 187–
230.
Shah, P. & Miyake, A. (1996). The separability of working memory resources for spatial
thinking and language processing: An individual differences approach. Journal of
Experimental Psychology. General, 125, 4–27.
Swatton, P. & Taylor, R. M. (1994). Pupil performance in graphical tasks and its
relationship to the ability to handle variables. British Educational Research Journal, 20,
227–243.
Vekiri, I. (2002). What is the value of graphical displays in learning? Educational
Psychology Review, 14(3), 261–312.
Wavering, M. J. (1989). Logical reasoning necessary to make line graphs. Journal of
Research in Science Teaching, 26, 373–379.
Woolnough, J. (2000). How do students learn to apply their mathematical knowledge to
interpret graphs in physics? Research in Science Education, 30, 259–267.

Maja Planinic, Ana Susac and Lana Ivanjek

Department of Physics, Faculty of Science
University of Zagreb
Bijenicka 32, 10000 Zagreb, Croatia
E-mail: maja@phy.hr

Ana Susac
E-mail: ana@phy.hr

Lana Ivanjek
E-mail: lana@phy.hr

Zeljka Milin-Sipus
Department of Mathematics, Faculty of Science
University of Zagreb
Bijenicka 30, 10000 Zagreb, Croatia
E-mail: milin@math.hr

Helena Katic
High school Slunj
Skolska 22, 47240 Slunj, Croatia
E-mail: helena.katic@ka.t-com.hr